Although overall obesity rates are declining in the US, severe obesity (BMI > 35 kg/m2) is still increasing; rates are as high as 36% in some US ethnic minority populations. Bariatric surgery is the most promising treatment for weight loss in the severely obese, resulting in much higher excess weight loss (75% vs. 11%) than traditional behavioral methods. The most frequently used bariatric procedures (bypass and gastric sleeve) have clear evidence for durable weight loss, however, even within the same procedure type, weight loss varies substantially. For example, 25% - 50% of RYGB patients regain some of their initial weight lost within 3 years. We propose a unique, mixed methods, prospective cohort study to understand why some patients experience better weight outcomes than others. The healthcare setting for this work, Kaiser Permanente Southern California (KPSC), performs ~3,500 bariatric surgeries annually; a larger target population for prospective research than any other single institution. Using socio-ecological, self-regulatory, and social cognitive theories fo the basis of our study design, we propose to collect electronic medical record data and self-report surveys in 1,800 patients before surgery and at 12, 24, and 36 months post-operatively to examine demographic, behavioral, psychosocial, and perceived environmental predictors of weight loss/regain. We will also purposively select cross-sectional samples from 12-, 24- and 36-month survey respondents who do or do not achieve successful weight loss to participate in interviews and focus groups. We will test the following two main study aims and one exploratory aim: Aim 1. Over a 3-year follow-up period, determine how demographic, behavioral, psychosocial, and perceived environmental pre-surgical factors, and post-surgical behavioral and psychosocial factors predict a) the weight loss trajectory and b) weight loss success (defined as achieving and maintaining > 50% excess weight loss); Aim 2. Understand the variability in weight loss using qualitative methods in cross-sectional samples of post-operative survey respondents and use these findings to inform mediational models addressing Aim 1; Exploratory Aim 3. Over a 3-year follow-up period, study the development of adverse psychosocial consequences and determine how they relate to a) overall weight loss and b) weight loss success. Our study is designed to provide results that will move the science forward in two key areas: 1) optimizing patient selection in real-world settings for bariatric procedures using a better understanding of pre-operative predictors; and 2) program development for maximizing the benefits from surgery for the greatest number of people, using the knowledge we gain from studying mediators of post-operative weight loss.